Fractal Summaries to Seed the Deliberations
2025-01-19
from Applying TalkToTheCity to Books
Fractal Summary to [Seed for deliberation
It's a little bit related to fractal summarization, and in the context of digital democracy, it's also about how to efficiently input knowledge from where it's needed in a situation where a lot of knowledge is needed.
In particular, the fractal summarization was very good in that it allowed me to summarize what was said in a conversation in a hierarchical manner while grounding it in the original speech.
So I think it would be useful to put the results of the fractal reservation of the conversation into TalkToTheCity, or some other digital democratic tool.
---
When I was using fractal summarization, I felt comfortable summarizing from videoconference discussions, probably because the amount of information in the "so-called summary" is too thin, with only "this topic was discussed", and the level of transcription is distracted by the mess caused by mistakes. I think there's value in a "summary that can be drilled down to the details as needed".
This is used for inputting information to people in the deliberation system, for example, a video of a dialogue between Mr. Yasuno and an expert is fractally summarized with "a slightly adjusted prompt for later use in the deliberation system," and the seed comments are then fed into Polis or some other platform where users can express their opinions and are seeded. The user's opinion and the opinion connected to the video are placed on the same two-dimensional map (in the same way that public opinion map places users and political parties in the same space).
2025-01-21
I was just sorting through it, and it doesn't really need to be a fractal summary.
Summary of the entire video
Detailed summary
TTTC's EXTRACTION to extract detailed opinions
I felt that two of the following were enough
https://gyazo.com/42d5027b671be8d114eeaf6370b40f59
---
This page is auto-translated from /nishio/フラクタル要約を熟議のシードにする using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.